Recent Submissions

  • Non-intrusive Microwave Sensor for In Situ Multiphase Flow Sensing in Oil Production

    Mansori, Hassan (2023-08-09) [Poster]
    Efficient oil production requires accurate flow rate and phase distribution monitoring to ensure safety and optimization. Traditional intrusive sensors have limitations, necessitating a novel approach. This work proposes a non-intrusive microwave T-resonator sensor to detect water and gas contents in oil production wells. The sensor's distinct behavior for different fluids is achieved using spiral resonators. The design is orientation insensitive, relatively inexpensive, and suitable for extreme well conditions. It eliminates the need for mixing oil and water, providing accurate measurements and promising a practical solution for multiphase flow monitoring.
  • Attentive Graph-based Relational Encoder (GRE) for Antonyms vs Synonyms Distinction

    Alshmrani, Maha Muhammed (2023-08-09) [Poster]
    Antonyms vs synonyms distinction is a core challenge in lexicon-semantic analysis and automated lexical resource construction. These pairs share a similar distributional context which makes it harder to distinguish them from each other. Leading research in this regard at007 tempts to capture the properties of the relation pairs, e.g., symmetry, transitivity, etc. How009 ever, the inability of existing research to appropriately model the relation-specific properties limits their end performance. In this paper, we propose GRE, i.e., Graph-based Relational Encoder that aims to capture and model the relation-specific properties of the antonyms and synonyms pairs in order to perform the classifi016 cation task in a performance-enhanced manner. Experimental evaluation using the benchmark datasets shows that GRE outperforms the existing research by a relative score of up to 1.8% in F1-measure.
  • Custom Data Set Recognition for Industrial Automation

    Kamal, Jana (2023-08-09) [Poster]
    This project is a combination of a Universal Robot manipulator controlled using ROS and MoveIt, a Zivid2 camera is attached to the manipulator to capture an image of what is infront of it to perform mug classification using AI by employing YOLOv5 with a custom dataset created using Roboflow. The system is able to identify whether a mug is intact or broken. This integration of robotics and AI allows for efficient and automated mug inspection and decision-making capabilities.
  • Object recognition using tactile sensor

    Azab, Meral (2023-08-09) [Poster]
    Tactile sensors have emerged as a promising technology for object identification, offering a distinct approach compared to traditional visual-based recognition systems. By relying on touch to detect and analyze physical properties, tactile sensors provide a more accurate and intuitive means of recognizing objects. This study aims to showcase the advantages of utilizing tactile sensors for object recognition. Through the development of a model, we demonstrate the effectiveness of tactile sensors in accurately recognizing objects. Our model achieved a high level of accuracy on the test set and successfully predicted the class label of sample images. Furthermore, this study highlights the potential of DIGIT tactile sensor 2D images for enhancing object recognition. Future explorations in this area aim to unlock this potential, paving the way for advancements in tactile-based perception systems.
  • Price Formation Using Mean Field Games

    Aljadhai, Khaled (2023-08-09) [Poster]
    Here, we consider price-formation model using mean-field game (MFG) framework. The model considers a large number of small players engage in trading commodity such as electricity. We present semi-explicit solution for the linear-quadratic model with time-varying preference included as a penalty. Then, we explore qualitative examples of the model.
  • Probing self-assembled monolayers on Au(111) substrates at low temperature

    musallam, arwa (2023-08-09) [Poster]
    To investigate the behaivor of SAMs on Au(111) at 150K using scanning tunneling microscopy.
  • Bayesian Deep Neural Networks

    AlMalawi, Maysoon (2023-08-09) [Poster]
    To investigate and develop methods for the fitting of Bayesian neural networks.
  • Characterization of Solid Microneedles Coated with Hydrogel for Interstitial Fluid Extraction

    bajunaid, Roba (2023-08-09) [Poster]
    Microneedle (MN) applications in healthcare vary widely, from transdermal drug and vaccine delivery to extraction of skin interstitial fluid (ISF). MNs are non-invasive due to their small size and design. They consist of tiny needle tips that penetrate the outermost layer of the skin without causing significant pain or damage. They vary in geometry (pyramidic, conical) typically having a height range of 50-900?m, and can be categorized into solid, hollow, and porous MNs. Herein, we developed a 3x3 solid MN array coated with a layer of hydrogel matrix for extraction of ISF due to their biocompatibility, water absorption and swelling behavioral properties. We experimented with different ratios of alginate to cross-linker to determine the optimal hydrogel that has the highest absorption volume and tested the mechanical strength of the MNs to ensure their capability of penetrating through skin. Results show that a ratio of 5:1 has high absorption properties, and a solid MN has an average fracture force of 0.1967 N.